The survival rate for patients with head and neck cancer (HNC) diagnosed with cervical lymph node (cLN) or distant metastasis is low. Genomic alterations in the HRAS oncogene are associated with advanced tumor stage and metastasis in HNC. Elucidation of the molecular mechanisms by which mutated HRAS (HRASmut) facilitates HNC metastasis could lead to improved treatment options for patients. Here, we examined metastasis driven by mutant HRAS in vitro and in vivo using HRASmut human HNC cell lines, patient-derived xenografts, and a novel HRASmut syngeneic model. Genetic and pharmacological manipulations indicated that HRASmut was sufficient to drive invasion in vitro and metastasis in vivo. Targeted proteomic analysis showed that HRASmut promoted AXL expression via suppressing the Hippo pathway and stabilizing YAP1 activity. Pharmacological blockade of HRAS signaling with the farnesyltransferase inhibitor tipifarnib activated the Hippo pathway and reduced the nuclear export of YAP1, thus suppressing YAP1-mediated AXL expression and metastasis. AXL was required for HRASmut cells to migrate and invade in vitro and to form regional cLN and lung metastases in vivo. In addition, AXL-depleted HRASmut tumors displayed reduced lymphatic and vascular angiogenesis in the primary tumor. Tipifarnib treatment also regulated AXL expression and attenuated VEGFA and VEGFC expression, thus regulating tumor-induced vascular formation and metastasis. Our results indicate that YAP1 and AXL are crucial factors for HRASmut-induced metastasis and that tipifarnib treatment can limit the metastasis of HNC tumors with HRAS mutations by enhancing YAP1 cytoplasmic sequestration and downregulating AXL expression.

Significance:

Mutant HRAS drives metastasis of head and neck cancer by switching off the Hippo pathway to activate the YAP1–AXL axis and to stimulate lymphovascular angiogenesis.

The main causes of mortality from head and neck cancer (HNC) are treatment failure and/or acquired resistance to therapy, which manifests either as local or regional recurrence with cervical lymph node (cLN) metastasis or recurrent disease with distant metastasis (1). Despite advances in chemoradiotherapy and immunotherapy protocols for the systemic treatment of patients with metastatic disease, the five-year survival rate of patients diagnosed with stage III or IV HNC remains low (1–3).

These poor survival rates provide the motivation to examine the molecular alterations associated with cLN and distant metastasis of HNC. Tumor metastasis is a complex and multistep process (4) that includes numerous cellular processes, such as apoptosis, invasion, and migration, which must be activated/used by tumor cells to establish metastatic disease (4, 5). Gene expression profiling of primary and metastatic HNC has identified several genes that are associated with metastatic processes, including the epithelial–mesenchymal transition (EMT), stemness, microvascular angiogenesis and lymphogenesis, extracellular matrix degradation, resistance to cell death, and metabolic reprograming, and thus contribute to regional and distant metastasis in HNC patients (6, 7). The HRAS mutation (HRASmut) is an important oncogenic driver found in 7% of human patients with papillomavirus–negative HNC (8) and in 15% of patients with recurrent metastatic disease (9). The major hot spot mutations in HRAS occur at codons 12, 13, and 61, being frequently detected in cancers of the salivary glands at Q61 and of the oral cavity at G12 (8). The frequency of mutations in HRAS increases after the acquisition of resistance to treatment with cetuximab, an EGFR inhibitor (10). Although in cancer types other than HNC, where HRAS is known to be central protein in metastases signaling cascades (11, 12), the molecular mechanisms by which HRASmut induces cLN and distant metastasis in HNC remains elusive.

In light of the high frequency of HRAS mutations in patients with HNC with advanced-stage disease, we sought to elucidate the intracellular machinery enhancing cLN and lung metastasis in these patients. Thus, to comprehensively study the mechanism of HRASmut on HNC progression and metastasis, we performed genetic and pharmacological manipulations to block HRAS function and its downstream signaling effectors in HRASmut cell lines, patient-derived xenografts (PDX), and a novel syngeneic HRASmut models.

Human cell lines

The HRAS-mutant human HNC cell lines, HN31 (G12D), UMSCC17B (Q61L), and CAL33 (Leibniz Institute DSMZ-German Collection of Microorganisms and Cell Cultures, GmbH, Germany), were maintained at 37°C in a humidified atmosphere of 5% CO2 in the relevant medium supplemented with 1% l-glutamine 200 mmol/L, 100 U each of penicillin and streptomycin, and 10% FBS. No Mycoplasmas were detected with the Mycoplasma Detection Kit (PCR; HyLabs).

Primary culture of mice tongue cells

Fresh tongue tissues from male KRT14 Cre mice were washed with Hanks' Balanced Salt Solution or PBS, cut into small pieces with sterile scissors, and treated with an enzyme mixture of collagenase (10 mg/mL; Thermo Fisher Scientific, catalog 17104019), hyaluronidase (1 mg/mL; MilliporeSigma, catalog H3506), and DNase (200 U/mL; Thermo Fisher Scientific, catalog18047019). The tissues were dissociated using the gentleMACS dissociator. Cells were filtered through a 70-μm cell strainer, centrifuged (300 × g, 5 min), and cultured in DMEM high glucose medium. The cultures were stained for the epithelial markers, cytokeratin 14 and E-cadherin, for verification of epithelial origin. All antibodies used in this study have been listed in Supplementary Table S1.

Transduction of primary cells

The concentrated pTomo HRASV12 shp53 (Supplementary Fig. S1A; ref. 13) lentivirus was used to transduce primary tongue cells in the presence of 8 μg/mL of Polybrene (Santa Cruz Biotechnology). After 48 hours of transduction, the medium was replaced with fresh medium without virus. Nontransduced cells eventually die after 10 days, and only transduced cells survive after 14 days. Only cells that were GFP+/RFP (green fluorescent protein+/red fluorescent protein) were considered to be transduced, because only epithelial cells with KRT14+ express Cre, and Cre helps to cleave RFP. Primary tongue cells transduced with this oncogenic vector expressed only GFP because the expression of HRAS is regulated by Cre.

Production of murine sgAXL CRISPR knockout cell lines

Lentiviruses were created by transfecting HEK293 cells with the viral plasmids psPAX2, pMD2.G, and lenticrispr V2 with sgRNAs, that is, a control sequence (shCT) or either of the 3 different sequences for the silencing of AXL expression (sgAXL1, sgAXL2, and sgAXL3, in-house constructed plasmids) using PEI as described in the Supplementary Materials and Methods. Viruses were collected after 48 to 72 hours and used for cell infection. Cells were seeded in 6-well plates (150,000 cells per well) and infected with the lentiviruses in the presence of Polybrene (MilliporeSigma, 5G-H9268). Cells were selected with puromycin (Thermo Fisher Scientific, A11138–03).

Protein array analysis

Cell lysates were analyzed for 84 oncogenes using the Proteome Profiler Human Oncogene Protein Array kit (R&D Systems) according to the manufacturer's instructions. In brief, cell lysates were prepared using lysis buffer from the relevant array kit, and protein concentration was measured by Bio-Rad protein assay. Cell lysates diluted in the array buffer were incubated with the ready-to-use pre-coated array membranes (blocked in blocking buffer provided with the kit) overnight at 4°C on a rocking platform shaker. The array membranes were washed three times (10 min) with washing buffer (provided with the kit) to remove any unbound proteins. The membranes were then incubated with the detection antibody cocktail for 1.5 hours at room temperature with shaking. The membranes were washed again (3 × 5 min each) with washing buffer and further incubated with diluted streptavidin—horseradish peroxidase for 30 minutes at room temperature with shaking. The excess buffer was removed, and the protein spots were detected by chemiluminescence 1 minute after exposure to the Chemi Reagent mix (from the kit). The arrays were visualized, and images were captured by Azure C300 Chemiluminescent Western Blot Imaging System (Azure Biosystems). The densitometric analysis of the protein array was performed using ImageJ software with the Protein Array Analyzer plugin (14).

Animal studies

C57/BL6 mice were purchased from Envigo. NOD.Cg-Prkdc Il2rg/SzJ (NSG) and KRT14 Cre mice were purchased from The Jackson Laboratory. Mice were housed in air-filtered laminar flow cabinets with a 12/12 h light/dark cycle and were fed food and water ad libitum. Mice were maintained and treated according to the institutional guidelines of Ben-Gurion University of the Negev. Animal experiments were approved by the Institutional Animal Care and Use Committee [IL.80–12–2015 (E), IL.29–05–2018 (E), IL.43–06–2019 (E), IL.44–06–2019 (E)].

PDX development and maintenance

HRASmut HNC PDX samples, namely, PDX1 (HN11–6062) and PDX2 (HN13–7313) harboring G12S and G13R mutations in HRAS, respectively, were first transplanted subcutaneously into the flanks of 6-week-old NSG mice. Upon successful tumor engraftment, tumors were expanded and retransplanted into other NSG mice. About 3 weeks after the second transplantation, each tumor was excised and a single-cell suspension was prepared. For the experiments, 3 × 106 cells were injected into the lips of NSG mice.

Spontaneous metastasis model

Orthotropic models were established by injecting 1 × 106 HRASmut human or murine tumor cells or their AXL modulated cells into the lips of C57/BL6 or NSG mice. Tumors were measured with a digital caliper twice a week, and tumor volumes were determined according to the formula: length × width2 × (π/6). At the end of the experiment, when the tumors had reached 500 mm3 in size, the animals were sacrificed by exposing them to CO2 inhalation. The tumors, draining cLNs, and lungs were harvested for investigation.

Drug efficacy studies

Animals were randomized into 2 or 4 groups of 5–6 mice per group, depending on the particular in vivo metastasis or drug efficacy experiment. When the tumors reached 100 mm3 in size, the animals were treated orally with tipifarnib (60 mg/kg/twice daily), or R428 (50 mg/kg/twice daily; provided by BerGenBio) in a vehicle of 0.5% carboxymethylcellulose (MilliporeSigma), as indicated in the Results section. Tumors were measured with a digital caliper twice a week, and tumor volumes were determined as described above. At the end of the experiment, animals were sacrificed by exposing them to CO2 inhalation, and the tumors, cLNs, and lungs were harvested for investigation. Measurements of tumor volumes are given either as average tumor volume ± SEM or as values normalized to initial volumes and presented as an averaged percentage of the initial volumes ± SEM. Harvested tissues were fixed in 4% paraformaldehyde (PFA) overnight and stored in 70% ethanol for further investigation. Serial sections of the lungs were stained with H&E for examining metastasis.

HUVEC maintenance

HUVECs were obtained from ScienCell Research Laboratories (No. 8000) and were grown in endothelial culture medium (No. 1001, ScienCell) containing 5% FBS (No. 0025), 1% endothelial cell growth supplement (No. 1052) and 1% penicillin/streptomycin solution (No. 0503) in 5% CO2 at 37°C. Cells were used in experiments between passages 2 and 5.

Conditioned media

HRASmut AXL knock down/knockout cells were allowed to grow to a confluence of 80%. After they reached confluence, the medium was replaced with a serum‐free DMEM. The conditioned medium (CM) was collected after 48 hours of incubation and centrifuged at 2,000 rpm for 15 minutes to remove cell debris. The supernatant, that is, the CM, was stored at −80°C until use.

For tipifarnib and R428 conditioned media. HRASmut human cell line HN31 and murine cells were allowed to grow to a confluence of 80%. After they reached confluence, the cells were treated with DMSO or tipifarnib (100 nmol/L/1 μmol/L)/R428 (2 μmol/L) for 24 hours. After 24 hours post treatment, the cells were washed with PBS thrice and supplemented with serum-free DMEM. The CM was collected after 48 hours of incubation and centrifuged at 2,000 rpm for 15 minutes to remove cell debris. The supernatant, that is, the CM, was stored at −80°C until use.

Statistical analysis

Statistical analysis was performed using GraphPad Prism software (version 9), and values are presented as means ± SEM. For comparisons between two groups, P values were calculated by an unpaired t test. One-way ANOVA was calculated using the Tukey multiple comparison test for experiments with more than two groups. P values of 0.05 (*), 0.01 (**), 0.001 (***), and 0.0001 (****) were considered statistically significant.

Data availability statement

The data generated in this study are available within the article and its Supplementary Data Files. Additional data are available on request from corresponding author. Detailed Materials and Methods are provided as Supplementary Materials and Methods.

Mutated HRAS enhances HNC cell metastasis via switching off the Hippo pathway and activating the YAP1–AXL axis

To explore whether HRAS mutations potentiate invasiveness of HNC cell line in vitro, we performed a gain-of-function study in which HRASmut was genetically incorporated (by using overexpression of the pTomo HRASV12 vector; ref. 13) into wild-type (WT) nonmetastatic HRAS Cre-expressing CAL33 (CAL33 Cre+) cells, followed by determination of tumor cells migration and invasion capability (Fig. 1A; Supplementary Fig. S1A and S1B). These in vitro studies showed that CAL33 Cre+ HRASV12+ cells exhibited enhanced migration and invasive potential compared with CAL33 Cre+ cells (Fig. 1B). In reciprocal studies, we examined whether silencing HRAS in HRASmut cells inhibits their invasive capability. We knocked down HRAS in the metastatic G12D-HRASmut HNC HN31 cell line by using RNA interference (RNAi; Fig. 1C; Supplementary Fig. S1C); in those experiments, a reduction in the invasion and migration capability of the knocked-down tumor cells was observed compared with cells transfected with scrambled/control RNAi (Fig. 1D). Notably, knockdown of HRAS did not affect the proliferation of HN31 tumor cells (Supplementary Fig. S1D).

Figure 1.

HRASmut regulates invasive potential in vitro via the modulation of AXL through YAP1 activation by switching off the Hippo pathway. A, Western blot confirming the ectopic expression of mutated HRASV12 in the nonmetastatic Cre-expressing CAL33 cell line. B, Bar graph showing enhanced migration and invasion ability of CAL33 Cre+ HRASV12+ cells in vitro. C, Western blot showing silencing of HRAS expression in an HRASmut HNC-HN31 cell line. D, Silencing of HRAS expression reduces migration and invasive potential of HN31 cells in vitro. E, Volcanic plot showing the effect of altered expression of oncogenic proteins on HRAS gain of function (CAL33) and loss of function (HN31). Proteins with a fold change of more than 1 or less than −1 with a P value less than 0.001 are shown in the plot. F, Venn diagram showing proteins altered during gain-of-function and loss-of-function manipulations of HRAS in CAL33 and HN31 cell lines. G, Western blot showing effect of enhanced AXL and YAP1 expression on ectopic expression of HRASV12 in CAL33 Cre-expressing cells. H, Western blot showing effect of AXL and YAP1 downregulation on silencing HRAS expression in HN31 cells. β-Actin was used as a loading control. I, Western blot confirming YAP1 silencing in HN31 cells, with a concomitant reduction in AXL expression. J, Bar graph showing that YAP1 silencing also reduces the migration and invasion potential of HN31 cells. K, Western blot showing the effect of the expression of various phosphorylated and total proteins in the Hippo pathway, namely, MST1/2, LATS1, YAP1, and TAZ, on silencing HRAS in HN31 cells. L, Western blot showing the downregulation of AXL and YAP1 in response to tipifarnib (1 μmol/L) treatment. M, Tipifarnib treatment of HN31 cells for 24 hours reduces the invasive capability of the cells. N, Western blot showing the expression of various phosphorylated and total proteins in the Hippo pathway, namely, MST1, LATS1, YAP1, and TAZ, upon treating HN31 cells with 1 μmol/L tipifarnib for 12 hours. O, Nuclear and cytoplasmic expressions of YAP1 in response to tipifarnib treatment. Actin and histone H3 were used as cytoplasmic and nuclear protein loading controls, respectively. Quantification of Western blots is provided as Supplementary Fig. S1B, S1C, and S1F–S1L. ****, P < 0.0001.

Figure 1.

HRASmut regulates invasive potential in vitro via the modulation of AXL through YAP1 activation by switching off the Hippo pathway. A, Western blot confirming the ectopic expression of mutated HRASV12 in the nonmetastatic Cre-expressing CAL33 cell line. B, Bar graph showing enhanced migration and invasion ability of CAL33 Cre+ HRASV12+ cells in vitro. C, Western blot showing silencing of HRAS expression in an HRASmut HNC-HN31 cell line. D, Silencing of HRAS expression reduces migration and invasive potential of HN31 cells in vitro. E, Volcanic plot showing the effect of altered expression of oncogenic proteins on HRAS gain of function (CAL33) and loss of function (HN31). Proteins with a fold change of more than 1 or less than −1 with a P value less than 0.001 are shown in the plot. F, Venn diagram showing proteins altered during gain-of-function and loss-of-function manipulations of HRAS in CAL33 and HN31 cell lines. G, Western blot showing effect of enhanced AXL and YAP1 expression on ectopic expression of HRASV12 in CAL33 Cre-expressing cells. H, Western blot showing effect of AXL and YAP1 downregulation on silencing HRAS expression in HN31 cells. β-Actin was used as a loading control. I, Western blot confirming YAP1 silencing in HN31 cells, with a concomitant reduction in AXL expression. J, Bar graph showing that YAP1 silencing also reduces the migration and invasion potential of HN31 cells. K, Western blot showing the effect of the expression of various phosphorylated and total proteins in the Hippo pathway, namely, MST1/2, LATS1, YAP1, and TAZ, on silencing HRAS in HN31 cells. L, Western blot showing the downregulation of AXL and YAP1 in response to tipifarnib (1 μmol/L) treatment. M, Tipifarnib treatment of HN31 cells for 24 hours reduces the invasive capability of the cells. N, Western blot showing the expression of various phosphorylated and total proteins in the Hippo pathway, namely, MST1, LATS1, YAP1, and TAZ, upon treating HN31 cells with 1 μmol/L tipifarnib for 12 hours. O, Nuclear and cytoplasmic expressions of YAP1 in response to tipifarnib treatment. Actin and histone H3 were used as cytoplasmic and nuclear protein loading controls, respectively. Quantification of Western blots is provided as Supplementary Fig. S1B, S1C, and S1F–S1L. ****, P < 0.0001.

Close modal

To identify the mechanism by which HRASmut induces enhanced migration/invasion in vitro, we used the human oncogene protein array for comparing (i) proteins expressed by nonmetastatic CAL33 Cre+ cells versus metastatic CAL33 Cre+ HRASV12+ cells and ii) proteins expressed by HN31 metastatic HRASmut cells silenced with HRAS-RNAi versus cells treated with scrambled-RNAi (Fig. 1E). By keeping a cutoff value of the fold change of more than 1 with a significance of P < 0.001, we identified 17 altered proteins in CAL33 Cre+ HRASV12+ and three proteins in HN31 cells that were altered under HRAS-silencing conditions. Among the proteins that were altered in response to HRAS modulation, the most common were AXL and PECAM-1: Their expression increased in CAL33 Cre+ HRASV12+ versus CAL33 Cre+ and declined in HRAS-RNAi HN31 cells versus HRAS-scrambled RNAi cells (Fig. 1F). AXL is a well-studied receptor tyrosine kinase (RTK) involved in the EMT and in metastasis (15), but the role of HRASmut in regulating AXL expression and in enhancing migration/invasion and metastasis has not been previously described. YAP1 is a known transcriptional activator of AXL expression (16) and remarkably an examination of The Cancer Genome Atlas (TCGA) HNC cohorts revealed a strong correlation between AXL and YAP1 expressions (Supplementary Fig. S1E). Thus, we hypothesized that HRASmut regulates the AXL expression through YAP1. To test this hypothesis, we investigated AXL and YAP1 expressions in CAL33 and HN31 cells with overexpression or knockdown of mutated HRAS. We found a concomitant modulation in the expression levels of YAP1 and AXL in these tumor cells with dysregulation of HRASmut (Fig. 1G and H; Supplementary Fig. S1F and S1G). We then explored whether YAP1 regulates AXL expression and whether this determines the migration and invasion potential of HRASmut HN31 cells. Silencing of YAP1 resulted in downregulation of AXL expression (Fig. 1I; Supplementary Fig. S1H) and a reduction in the ability of the cells to migrate and invade (Fig. 1J). As HRASmut can regulate the Hippo pathway and YAP1 expression and activity (17), we analyzed the activation status of the Hippo pathway in HRASmut-HN31 cells treated with siHRAS and siControl. We found that silencing HRAS expression activated the Hippo pathway, as indicated by an increase of pLATS1 and pMST1/2, the major activators of the Hippo pathway. Moreover, the reduction of YAP1 expression was associated with increased phosphorylation of YAP1 at serine 109, 127, and 397, which affected the cytoplasmic retention of YAP1 and its subsequent degradation (Fig. 1K; Supplementary Fig. S1I; ref. 18).

We then explored whether pharmacological inhibition of HRAS signaling could induce the same phenotype as that obtained by knocking down HRAS. Blocking HRAS signaling in HRASmut-HN31 cells with tipifarnib, a farnesyl transferase inhibitor, reduced AXL and YAP1 expression (Fig. 1L; Supplementary Fig. S1J) and lowered the invasive potential of these cells (Fig. 1M). Finally, we showed that tipifarnib treatment activated the Hippo pathway and increased the phosphorylation of YAP1 at positions serine 109, 127 and 397 (Fig. 1N; Supplementary Fig. S1K) and also reduced the nuclear expression of YAP1 (Fig. 1O; Supplementary Fig. S1L). These findings indicate that HRASmut regulates AXL expression by switching off the Hippo pathway, increasing nuclear export and hence activity of YAP1, thus increasing the invasive potential of cancer cells.

Metastasis potential of HRASmut cells is associated with AXL and YAP1 overexpressions

To test the spontaneous metastasis of HRASmut tumors, we injected CAL33 Cre+ or CAL33 Cre+ HRASV12+ cells into the lips of immunodeficient NSG mice, measured primary tumor growth over time (Fig. 2A), and quantified the cLN and lung metastasis when the tumors reached approximately 500 mm3 in size. CAL33 Cre+ HRASV12+ cells exhibited enhanced metastatic potential, as shown by the enlarged cLNs and the greater number of lung micro-metastasis foci in CAL33 Cre+ HRASV12+-injected NSG mice versus CAL33 Cre+-injected mice (Fig. 2B). IHC staining for tumor cells confirmed the accumulation of tumor cells in the cLNs and lungs of the mice (Fig. 2C; Supplementary Fig. S2A). To show that our findings were reproducible in independent HRASmut HNC models, we performed spontaneous metastasis studies using cell line–derived xenografts (CDX) and PDXs possessing HRAS mutations. To develop CDXs, we injected HRASmut UMSCC17B (Q61L) or HN31 cells orthotopically into the lips of NSG mice. The UMSCC17B cells grew more rapidly (Fig. 2D), but the mice exhibited a low number of enlarged cLNs, and only 4 of the 12 mice showed micro-metastases. In contrast, growth of the HN31 cells was slower, but these cells exhibited enhanced metastatic capacity, reaching the cLNs and the lungs more efficiently with macrometastases (Fig. 2E; Supplementary Fig. S2B). Western blot analysis of the cell lines showed that the metastatic HN31 cells expressed higher levels of YAP1 and AXL than the weakly metastatic UMSCC17B cells (Fig. 2F; Supplementary Fig. S2C). To study the metastasis of HRASmut PDXs, we used PDX1 and PDX2, the two PDXs described previously in Materials and Methods. Briefly, we injected 3 × 106 single-cell suspensions of PDXs into the lips of NSG mice and allowed tumors to form. When the tumors reached approximately 500 mm3 in size, the mice were sacrificed, and the number of enlarged cLNs and the number of metastatic foci in the lungs of each tumor-bearing mouse were counted. PDX1 (HN11–6062) showed low metastatic potential with no cLN metastases, but lung micro-metastases were evident in 2 of 12 mice. In contrast, the metastatic potential of PDX2 (HN13–7313) was significantly higher as all 12 mice exhibited cLN metastases, and 5 of the 12 mice developed lung metastases (Fig. 2G and H; Supplementary Fig. S2D). Western blot analysis of the PDXs showed that the metastatic PDX2 expressed higher levels of AXL and YAP1 than the less-metastatic PDX1 (Fig. 2I; Supplementary Fig. S2E). These findings provide support for the association of AXL and YAP1 expressions with the enhanced metastasis potential of HRASmut HNC models.

Figure 2.

The HRAS–YAP1–AXL axis determines the metastatic potential of human HNC cells in vivo.A and B,In vivo analysis of tumor growth kinetics (A) and number of enlarged lymph nodes and lung metastasis foci (B) in mice (n = 12) bearing CAL33 Cre+ or CAL33 Cre+ HRASV12+ tumors; the figures show the enhanced metastasis potential on the incorporation of HRAS mutation. C, Representative IHC images of the lung metastases (stained for pan keratin) of CAL33 Cre+ and CAL33 Cre+ HRASV12+. Inset magnification, ×20. D and E,In vivo analysis of tumor growth kinetics (D) and number of enlarged lymph nodes and lung metastasis foci (E) in mice (n = 12) injected with HRASmut HNC cell lines—UMSCC17B or HN31; the figures show the high metastasis potential in HN31 cells. F, Western blot showing the expression of AXL and YAP1 in UMSCC17B and HN31 cell lines. β-Actin was used as a loading control. G and H,In vivo representation of tumor growth kinetics (G) and number of enlarged lymph nodes and lung metastasis foci (H) in mice (n = 12) bearing HRAS mutant HNC PDXs (PDX1 and PDX2) after injection with single-cell suspensions into the lip of the NSG mice. I, Western blot showing the expression of AXL and YAP1 in PDXs. J and K,In vivo analysis of relative tumor growth (J) and the numbers of enlarged lymph nodes and lung metastasis foci (K) in HN31 tumor-bearing mice (n = 5–6) treated with tipifarnib (60 mg/kg/twice daily); the findings show a reduced tumor and metastasis burden. L, Representative images of lung macrometastasis; lungs were isolated from vehicle-and tipifarnib-treated HN31 tumor-bearing mice (n = 5). Quantification of Western blots are provided as Supplementary Fig. S2C and S2E. *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001.

Figure 2.

The HRAS–YAP1–AXL axis determines the metastatic potential of human HNC cells in vivo.A and B,In vivo analysis of tumor growth kinetics (A) and number of enlarged lymph nodes and lung metastasis foci (B) in mice (n = 12) bearing CAL33 Cre+ or CAL33 Cre+ HRASV12+ tumors; the figures show the enhanced metastasis potential on the incorporation of HRAS mutation. C, Representative IHC images of the lung metastases (stained for pan keratin) of CAL33 Cre+ and CAL33 Cre+ HRASV12+. Inset magnification, ×20. D and E,In vivo analysis of tumor growth kinetics (D) and number of enlarged lymph nodes and lung metastasis foci (E) in mice (n = 12) injected with HRASmut HNC cell lines—UMSCC17B or HN31; the figures show the high metastasis potential in HN31 cells. F, Western blot showing the expression of AXL and YAP1 in UMSCC17B and HN31 cell lines. β-Actin was used as a loading control. G and H,In vivo representation of tumor growth kinetics (G) and number of enlarged lymph nodes and lung metastasis foci (H) in mice (n = 12) bearing HRAS mutant HNC PDXs (PDX1 and PDX2) after injection with single-cell suspensions into the lip of the NSG mice. I, Western blot showing the expression of AXL and YAP1 in PDXs. J and K,In vivo analysis of relative tumor growth (J) and the numbers of enlarged lymph nodes and lung metastasis foci (K) in HN31 tumor-bearing mice (n = 5–6) treated with tipifarnib (60 mg/kg/twice daily); the findings show a reduced tumor and metastasis burden. L, Representative images of lung macrometastasis; lungs were isolated from vehicle-and tipifarnib-treated HN31 tumor-bearing mice (n = 5). Quantification of Western blots are provided as Supplementary Fig. S2C and S2E. *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001.

Close modal

Because treatment of HRASmut cells with tipifarnib reduced YAP and AXL expressions and reduced cell migration and invasion in vitro, we posited that tipifarnib treatment of mice bearing metastatic HRASmut-HN31 tumors would also attenuate spontaneous metastasis into the cLNs and lungs of the mice. Indeed, we found that daily treatment of HN31 tumor-bearing mice with tipifarnib resulted in growth retardation of the primary tumor (Fig. 2J) accompanied by a reduction of enlarged cLNs and lung macrometastasis (Fig. 2K and L).

AXL expression potentiates metastasis of HRASmut HNC cells

We next set out to determine whether AXL expression is required for metastasis of HRASmut HNC. To this end, we initially knocked down AXL expression in the metastatic HN31 cell line (Fig. 3A; Supplementary Fig. S3A) and then tested the ability of the tumor cells to migrate/invade in vitro and metastasize in vivo. The in vitro experiment showed that HRASmut-HN31 cells with AXL knockdown (shAXL) were less invasive than HN31 control cells (shCT; Fig. 3B; Supplementary Fig. S3B). We further showed that AXL activation by its ligand Gas6 is required for the migration of HN31 tumor cells and that the upregulation of Gas6 in shAXL–HN31 tumor cells was not sufficient to facilitate cell migration (Fig. 3C; Supplementary Fig. S3C). In vivo experiments revealed that AXL knockdown enhanced the primary growth of HRASmut-HN31 xenografts (Fig. 3D), but these cells lost the ability to promote metastasis to the cLNs and lungs (Fig. 3E; Supplementary Fig. S3D). Transwell migration assay showed that pharmacological inhibition of AXL with R428, a known potent and selective inhibitor of AXL, reduced the ability of the HN31 cells to migrate and invade (Fig. 3F). This reduction in invasion and migration was associated with a plasticity change of the mesenchymal-like phenotype of HN31 from N-cadherin high and E-cadherin low cells to an epithelial-like phenotype with high E-cadherin and lower N-cadherin, but with no inhibition of the MAPK or the PI3K/AKT pathway (Supplementary Fig. S3E). Finally, in HN31 tumor-bearing mice that were treated twice daily with R428, growth of the primary tumor was not delayed and there was no significant impact on tumor volumes (Fig. 3G), but there was a reduction in cLN and lung metastasis (Fig. 3H). Similar results were also obtained in PDX2 tumor-bearing mice treated with R428 (Fig. 3I and J). These results suggest that targeting AXL signaling impedes the metastasis capability of HRASmut HNC tumor cells.

Figure 3.

Genetic/pharmacological dysregulation of AXL signaling modulates the metastatic potential of human HRASmut HNC cells. A, Western blot and qPCR confirming the AXL knockdown of HN31 cells. B, Reduced migration capability of HN31 cells. C, Reduced migration capability of HN31 cells upon treatment with anti-Gas6. D and E,In vivo analysis of tumor growth kinetics (D) and number of enlarged lymph nodes and lung metastasis foci (E) in AXL knockdown HN31 tumor-bearing mice (n = 5). F, Bar graphs showing that pharmacological inhibition of AXL signaling by R428 (2 μmol/L) reduces the invasive potential of HN31 cells in vitro.G and H,In vivo analysis of tumor growth kinetics (G) and number of enlarged lymph nodes and lung metastasis foci (H) in HN31 tumor-bearing mice (n = 5) treated with R428 (50 mg/kg/twice daily). I and J,In vivo analysis of tumor growth kinetics (I) and number of enlarged lymph nodes and lung metastasis foci (n = 5; J) in PDX2 tumor-bearing mice treated with R428 (50 mg/kg/twice daily). K, AXL overexpression (AXL OE) in weakly metastatic UMSCC17B cell line enhances invasive potential in vitro compared with vector control. L and M,In vivo analysis of tumor growth kinetics (L) and number of enlarged lymph nodes and lung metastasis foci (M) in AXL-overexpressing UMSCC17B cells injected mice (n = 5). Quantification of Western blots is provided as Supplementary Fig. S3A. *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001; ns, not significant.

Figure 3.

Genetic/pharmacological dysregulation of AXL signaling modulates the metastatic potential of human HRASmut HNC cells. A, Western blot and qPCR confirming the AXL knockdown of HN31 cells. B, Reduced migration capability of HN31 cells. C, Reduced migration capability of HN31 cells upon treatment with anti-Gas6. D and E,In vivo analysis of tumor growth kinetics (D) and number of enlarged lymph nodes and lung metastasis foci (E) in AXL knockdown HN31 tumor-bearing mice (n = 5). F, Bar graphs showing that pharmacological inhibition of AXL signaling by R428 (2 μmol/L) reduces the invasive potential of HN31 cells in vitro.G and H,In vivo analysis of tumor growth kinetics (G) and number of enlarged lymph nodes and lung metastasis foci (H) in HN31 tumor-bearing mice (n = 5) treated with R428 (50 mg/kg/twice daily). I and J,In vivo analysis of tumor growth kinetics (I) and number of enlarged lymph nodes and lung metastasis foci (n = 5; J) in PDX2 tumor-bearing mice treated with R428 (50 mg/kg/twice daily). K, AXL overexpression (AXL OE) in weakly metastatic UMSCC17B cell line enhances invasive potential in vitro compared with vector control. L and M,In vivo analysis of tumor growth kinetics (L) and number of enlarged lymph nodes and lung metastasis foci (M) in AXL-overexpressing UMSCC17B cells injected mice (n = 5). Quantification of Western blots is provided as Supplementary Fig. S3A. *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001; ns, not significant.

Close modal

To explore whether AXL has the ability to enhance metastasis in HRASmut HNC cell lines, we overexpressed (OE) AXL in the weakly metastatic UMSCC17B cells (Supplementary Fig. S3F) and determined their invasive potential in vitro and in mice. The in vitro experiment showed that overexpression of AXL enhanced the migration and invasion of UMSCC17B cells (Fig. 3K). The in vivo experiment in NSG mice showed similar growth kinetics for UMSCC17B-OE-AXL cells and UMSCC17B cells (Fig. 3L), but UMSCC17B-OE-AXL tumor-bearing mice exhibited an increased number of enlarged cLNs. Five of six mice bearing UMSCC17B-OE-AXL tumors showed lung metastasis with multiple foci compared with AXL-null vector-UMSCC17B tumors (Fig. 3M; Supplementary Fig. S3G). Thus, overexpression of AXL potentiates the metastasis capability of HRASmut HNC tumor cells.

Depleting AXL in HRASmut tumors reduces blood vessel formation and lymphangiogenesis by downregulation of the expression of VEGF A and C

To get further mechanistic insight on metastasis cascade, we hypothesized that HRASmut HNC cells that spread from the primary lesion to the cLNs and distant organs must first invade the vascular system and survive anoxic conditions. Analysis of the TCGA database supports this hypothesis as patients with HNC with mutated or amplified HRAS express high levels of lymphangiogenesis and angiogenesis signatures (Supplementary Fig. S4A). Moreover, comparing gene expression of metastatic HNC (POG570; ref. 19) and non-metastatic HNC (TCGA) also show HRAS overexpression, lymphangiogenesis, and angiogenesis activation in metastatic disease (Supplementary Fig. S4B). To test this hypothesis in our cancer models, we determined whether HRASmut tumors with AXL overexpression are enriched with lymphatic and blood vessels compared with tumors with low expression of AXL. Indeed, staining shCT and shAXL HN31 tumors with LYVE-1 (marker for lymphangiogenesis) and CD31 (marker for angiogenesis) revealed a lower number of stained cells in shAXL tumors (Fig. 4A). To explore the mechanism underlying such differences, we searched the TCGA dataset for VEGFA or VEGFC, both important angiogenesis factors involved in vessel formation in HNC patients (20–22), which may display a correlation between mRNA levels of AXL. The results revealed that AXL expression is strongly associated with VEGFC (Supplementary Fig. S4C), and we therefore posited that AXL-intact cells express and secrete VEGFs that support the angiogenesis process. The CM derived from AXL-intact shCT HN31 cells enhanced tube formation of HUVECs in vitro, but the CM derived from shAXL tumor cells showed reduced ability to induce tube formation of HUVEC (Fig. 4B). The base-line mRNA expression of VEGFA and VEGFC in high- and low-AXL-expressing tumor cells (shCT and shAXL) showed that cells with downregulation of AXL express lower mRNA levels of VEGFC and VEGFA compared with AXL-intact cells (Fig. 4C). Reduction of VEGFA and VEGFC was also observed in HN31 tumor cells treated with R428, and the CM of R428-treated cells induced less tube formation than the CM obtained from DMSO-treated cells (Fig. 4D and E). Moreover, neutralization of VEGFA using Avastin in the CM of shCT cells was sufficient to reduce tube formation (Fig. 4F), and supplementation of recombinant VEGFA in the CM of shAXL cells enhanced the formation of tubes by HUVEC (Fig. 4G).

Figure 4.

AXL knockdown in HRASmut HNC attenuates vascular/lymphatic angiogenesis by downregulating VEGFA and VEGFC expression. A, Immunostaining with CD31 and LYVE-1 (markers for angiogenesis and lymphangiogenesis, respectively) in primary tumors of AXL knockdown HN31; the micrographs show minimal angiogenesis and lymphangiogenesis upon AXL downregulation. The percentage of positive cells per field for each stain is provided on the right. B, Representative microscopic images of HUVEC tube formation with conditioned media from AXL knockdown HN31. Bar graphs representing the quantification of total mesh area on tube formation under each condition are shown on the right. C, Fold change in mRNA expression of VEGFA and VEGFC upon AXL knockdown in HN31 cells. D, Representative microscopic images of HUVEC tube formation using conditioned media from HN31 cells treated with R428 (2 μmol/L) or DMSO for 24 hours. E, Fold change in mRNA expression of VEGFA and VEGFC upon treatment of HN31 cells with R428 (2 μmol/L). F and G, Representative microscopic images of HUVEC tube formation using conditioned media from shCT HN31 cells treated with Avastin (10 μg/mL; F) and in conditioned media from shAXL HN31 cells treated with VEGFA (20 ng/mL; G). Quantification of total mesh area on tube formation under each condition are shown on the right. H, Fold change in mRNA expression of VEGFA and VEGFC upon treatment of HN31 cells with tipifarnib (1 μmol/L). I, Representative microscopic images of HUVEC tube formation using conditioned media from HN31 cells treated with tipifarnib (1 μmol/L) or DMSO for 24 hours. **, P < 0.01; ****, P < 0.0001; ns, not significant.

Figure 4.

AXL knockdown in HRASmut HNC attenuates vascular/lymphatic angiogenesis by downregulating VEGFA and VEGFC expression. A, Immunostaining with CD31 and LYVE-1 (markers for angiogenesis and lymphangiogenesis, respectively) in primary tumors of AXL knockdown HN31; the micrographs show minimal angiogenesis and lymphangiogenesis upon AXL downregulation. The percentage of positive cells per field for each stain is provided on the right. B, Representative microscopic images of HUVEC tube formation with conditioned media from AXL knockdown HN31. Bar graphs representing the quantification of total mesh area on tube formation under each condition are shown on the right. C, Fold change in mRNA expression of VEGFA and VEGFC upon AXL knockdown in HN31 cells. D, Representative microscopic images of HUVEC tube formation using conditioned media from HN31 cells treated with R428 (2 μmol/L) or DMSO for 24 hours. E, Fold change in mRNA expression of VEGFA and VEGFC upon treatment of HN31 cells with R428 (2 μmol/L). F and G, Representative microscopic images of HUVEC tube formation using conditioned media from shCT HN31 cells treated with Avastin (10 μg/mL; F) and in conditioned media from shAXL HN31 cells treated with VEGFA (20 ng/mL; G). Quantification of total mesh area on tube formation under each condition are shown on the right. H, Fold change in mRNA expression of VEGFA and VEGFC upon treatment of HN31 cells with tipifarnib (1 μmol/L). I, Representative microscopic images of HUVEC tube formation using conditioned media from HN31 cells treated with tipifarnib (1 μmol/L) or DMSO for 24 hours. **, P < 0.01; ****, P < 0.0001; ns, not significant.

Close modal

Because tipifarnib treatment is known to reduce tumor angiogenesis in HRASmut tumors (23), we explored whether tipifarnib treatment reduces VEGFA and VEGFC expression in both human and murine HRAS-mutant lines. qPCR analysis showed that treatment of HRASmut cells with tipifarnib reduced VEGFA and VEGFC mRNA expression in both cell lines (Fig. 4H). Finally, in a tube-formation assay, we exposed HUVECs to conditioned media collected from tumor cells subjected to 24 hours of pre-treatment with tipifarnib (or DMSO as control). The CM of the HN31 cells pre-treated with tipifarnib supported low tube formation compared with the CM of DMSO-pretreated HRASmut cells (Fig. 4I). These results suggest that AXL expression in HRASmut tumor cells augments the neovascularization process thus facilitating to the process of metastasis.

AXL is essential for metastasis in a syngeneic HRASmut HNC model

Metastasis is a multistage process, affected by multiple cell types in the microenvironment. To validate our findings in a physiologically relevant system, we generated an HRAS mutant cell line by lentiviral infection of primary tongue cells isolated from KRT14 Cre mice, and established a novel immune competent murine model of HNC (Fig. 5A; Supplementary Fig. S5A). The novel murine HNC cell line was designated HRASV12 shp53 EpT (HRASV12 shp53 epithelial tongue; Supplementary Fig. S5B and S5C). We confirmed the expression of HRAS and p53 knockdown by Western blotting and showed that the infection of normal tongue cells with the lentivirus induced an EMT phenotype with high expression of N-cadherin, AXL, and YAP1 (Fig. 5A; Supplementary Fig. S5D). The transformed HRASV12 shp53 EpT cells showed a tumorigenic phenotype compared with normal tongue cells, as was evident from the uncontrolled cell proliferation and the ability to form colonies and grow in anchorage-independent 3D cultures (Supplementary Fig. S5F and S5G). Similarly, compared with normal epithelial cells, HRASV12 shp53 EpT cells exhibited enhanced metastatic activity, with the ability to migrate and invade, as shown in Transwell assays (Supplementary Fig. S5H). The transform HRASV12 shp53 EpT cells displayed an inactivation of the Hippo pathway, as indicated by the decrease of pLATS1 and pMST1/2, the major activators of the Hippo pathway. Moreover, the increase of YAP1 expression was associated with decreased phosphorylation of YAP1 at serine 109 and 397 (Supplementary Fig. S5I). In vivo injection of HRASV12 shp53 EpT cells into the lips of WT C57/B6 mice resulted in measurable tumor growth after seven days (Supplementary Fig. S5J). Histological examination revealed poorly differentiated tumors with a spindle morphology and mesenchymal-type phenotype attributes, indicating an aggressive type of HNC (Supplementary Fig. S5K). To further study the spontaneous metastasis potential of HRASV12 shp53 EpT cells, we injected cells into the lips of NSG or WT mice. Interestingly, although the tumors exhibited the same growth kinetics in WT and NSG mice (Fig. 5B), the metastatic burden in the lungs was higher in the NSG mice than that in the WT mice (Fig. 5C). In contrast, the metastatic burden in the cLNs was enhanced in WT mice (Fig. 5C). cLN metastasis in WT mice was verified by flow cytometry and histology (Supplementary Fig. S5L and S5M). Importantly, all NSG mice injected with HRASV12 shp53 EpT cells exhibited lung macro-metastases, whereas in the WT mice, mainly micrometastases were detected in half of the mice, as confirmed by staining with GFP (Supplementary Fig. S5M). These results indicate that HRASV12 shp53 EpT cells induce a local immunosuppressive environment that facilitates rapid tumor growth, but upon tumor cell dissemination from the primary lesion the antitumor immunity machinery potentially eradicated such circulating tumor cells.

Figure 5.

Metastatic potential of murine HRASmut HNC is modulated by AXL. A, Top, schematic representation of the generation of HRAS mutant murine cells from primary KRT14 Cre+ tongue epithelial cells. Bottom, Western blot showing the expression of E-cadherin, N-cadherin, AXL, YAP1, p53, HRASV12, and β-actin in primary tongue epithelial cells (10 EpT) and HRASV12 lentiviral-transduced tongue epithelial cells (HRASV12 shp53 EpT). B and C,In vivo analysis of tumor growth kinetics (B) and number of enlarged lymph nodes and lung metastasis foci (C) of HRASV12 shp53 EpT in WT and NSG mice (n = 6). D and E,In vivo analysis of tumor growth kinetics, number of enlarged lymph nodes and lung metastasis foci in AXL knockout HRASV12 shp53 EpT tumor-bearing WT mice (n = 6; D) and AXL knockout HRASV12 shp53 EpT tumor-bearing NSG mice (n = 6; E) showing a reduced metastatic burden. F, Immunostaining with CD31 and LYVE-1 in primary tumors of AXL knockout HRASV12 shp53 EpT showing minimal angiogenesis and lymphangiogenesis upon AXL downregulation. The percentage of positive cells per field for each staining is shown on the right. G, Representative microscopic images of HUVEC tube formation with conditioned media from AXL knockout HRASV12 shp53 EpT. Bar graphs representing the quantification of total mesh area on tube formation under each condition are shown on the right. H, Fold change in mRNA expression of VEGFA and VEGFC upon AXL knockout in HRASV12 shp53 EpT cells. Quantification of Western blots are provided as Supplementary Fig. S5D. *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001; ns, not significant. (A, Created with BioRender.com.)

Figure 5.

Metastatic potential of murine HRASmut HNC is modulated by AXL. A, Top, schematic representation of the generation of HRAS mutant murine cells from primary KRT14 Cre+ tongue epithelial cells. Bottom, Western blot showing the expression of E-cadherin, N-cadherin, AXL, YAP1, p53, HRASV12, and β-actin in primary tongue epithelial cells (10 EpT) and HRASV12 lentiviral-transduced tongue epithelial cells (HRASV12 shp53 EpT). B and C,In vivo analysis of tumor growth kinetics (B) and number of enlarged lymph nodes and lung metastasis foci (C) of HRASV12 shp53 EpT in WT and NSG mice (n = 6). D and E,In vivo analysis of tumor growth kinetics, number of enlarged lymph nodes and lung metastasis foci in AXL knockout HRASV12 shp53 EpT tumor-bearing WT mice (n = 6; D) and AXL knockout HRASV12 shp53 EpT tumor-bearing NSG mice (n = 6; E) showing a reduced metastatic burden. F, Immunostaining with CD31 and LYVE-1 in primary tumors of AXL knockout HRASV12 shp53 EpT showing minimal angiogenesis and lymphangiogenesis upon AXL downregulation. The percentage of positive cells per field for each staining is shown on the right. G, Representative microscopic images of HUVEC tube formation with conditioned media from AXL knockout HRASV12 shp53 EpT. Bar graphs representing the quantification of total mesh area on tube formation under each condition are shown on the right. H, Fold change in mRNA expression of VEGFA and VEGFC upon AXL knockout in HRASV12 shp53 EpT cells. Quantification of Western blots are provided as Supplementary Fig. S5D. *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001; ns, not significant. (A, Created with BioRender.com.)

Close modal

We then went on to explore the role of AXL in the syngeneic HRASV12 shp53 EpT model by generating three AXL-knockout clones with CRISPR-Cas9. The knockout of AXL was confirmed by flow cytometry and Western blotting (Supplementary Fig. S5N and S5O). Subsequent studies with two of these clones, sgAXL1 and sgAXL2, showed that knockout of AXL in HRASV12 shp53 EpT cells resulted in a reduction of invasion and migration potential in vitro (Supplementary Fig. S5N) and in lung micro-metastasis in WT mice and macro-metastasis in NSG mice (Fig. 5D and E).

We then explored whether the HRASV12 shp53 EpT AXL knockout tumors display reduced angiogenesis, as observed in human HRASmut AXL knockdown HNC. Staining sgCT and sgAXL HRASV12 shp53 EpT tumors with LYVE-1 and CD31 revealed a lower number of stained cells in sgAXL tumors (Fig. 5F). Similar to the results obtained for the human HRASmut line, the CM derived from sgAXL tumor cells showed reduced ability to induce tube formation of HUVECs (Fig. 5G) and vessel formation in vivo (Supplementary Fig. S5P). qPCR analysis of the AXL knockout HRASV12 shp53 EpT cells also confirmed that downregulation of AXL leads to reduced VEGFA and VEGFC expression (Fig. 5H). These findings reiterate the potential role of AXL in HRASmut HNC cells in governing metastasis.

Pharmacological inhibition of AXL/HRAS signaling attenuates metastasis in a syngeneic model

We next explored the effect of AXL and HRAS inhibition on metastasis in HRASV12 shp53 EpT cells. Pharmacological inhibition of AXL activity by R428 blocked the metastatic capability of HRASV12 shp53 EpT cells in vitro (Fig. 6A). In vivo, treatment of tumor-bearing mice with R428 did not induce tumor growth arrest but reduced metastasis burden in both WT and NSG mice (Fig. 6B and C; Supplementary Fig. S6A). These results provide further support for our premise that AXL signaling is required for metastasis but not for primary tumor growth.

Figure 6.

Inhibition of HRAS/AXL signaling diminishes the metastatic potential of murine HRASmut cells. A, Pharmacological inhibition of AXL signaling with R428 reduced the migration and invasion of HRASV12 shp53 EpT. B and C,In vivo analysis of tumor growth kinetics and numbers of enlarged lymph nodes and lung metastasis foci in HRASV12 shp53 EpT tumor upon treatment with R428 (50 mg/kg/twice daily) in WT mice (n = 5; B) and NSG mice (n = 5; C) showing R428 treatment reduced the metastatic burden without affecting the tumor growth in both WT and NSG mice. D, Pharmacological inhibition of HRAS signaling with tipifarnib reduced the migration and invasion of HRASV12 shp53 EpT. E and F,In vivo analysis of tumor growth kinetics and numbers of enlarged lymph nodes and lung metastasis foci in HRASV12 shp53 EpT tumor-bearing WT (E) and NSG (F) mice (n = 6) treated with tipifarnib (60 mg/kg/twice daily) showing tumor growth delay and reduced metastatic burden in both WT and NSG mice. *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001; ns, not significant.

Figure 6.

Inhibition of HRAS/AXL signaling diminishes the metastatic potential of murine HRASmut cells. A, Pharmacological inhibition of AXL signaling with R428 reduced the migration and invasion of HRASV12 shp53 EpT. B and C,In vivo analysis of tumor growth kinetics and numbers of enlarged lymph nodes and lung metastasis foci in HRASV12 shp53 EpT tumor upon treatment with R428 (50 mg/kg/twice daily) in WT mice (n = 5; B) and NSG mice (n = 5; C) showing R428 treatment reduced the metastatic burden without affecting the tumor growth in both WT and NSG mice. D, Pharmacological inhibition of HRAS signaling with tipifarnib reduced the migration and invasion of HRASV12 shp53 EpT. E and F,In vivo analysis of tumor growth kinetics and numbers of enlarged lymph nodes and lung metastasis foci in HRASV12 shp53 EpT tumor-bearing WT (E) and NSG (F) mice (n = 6) treated with tipifarnib (60 mg/kg/twice daily) showing tumor growth delay and reduced metastatic burden in both WT and NSG mice. *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001; ns, not significant.

Close modal

To test the effect of tipifarnib on HRASV12 shp53 EpT cells, we initially explored the sensitivity of the cells to this agent and showed that the IC50 value is 1.1 μmol/L (Supplementary Fig. S6B). We next step investigated whether HRAS signaling in HRASV12 shp53 EpT cells determines AXL and YAP1 expressions and metastasis potential. Western blot analysis showed that treatment of the tumor cells with tipifarnib reduced both AXL and YAP1 expressions (Supplementary Fig. S6C and S6D). To explore the effect of tipifarnib on invasion and migration in vitro, we treated the HRASV12 shp53 EpT cells with 1 μmol/L tipifarnib and quantified the number of cells that migrated/invaded through the Transwell membrane. The Transwell migration/invasion assay showed that tipifarnib treatment reduced the ability of the HRASV12 shp53 EpT cells to migrate and invade (Fig. 6D). Finally, we tested the efficacy of tipifarnib in vivo in WT and NSG mice. Tipifarnib treatment of mice bearing HRASV12 shp53 EpT tumors inhibited tumor growth, and markedly reduced cLN and lung metastasis in both WT and NSG mice in both models (Fig. 6E and 6F).

Taken together, our findings support a novel molecular mechanism of metastasis in HNC-harboring HRAS mutation (Fig. 7). Specifically, HRASmut stabilizes YAP1 activity to increase AXL expression by turning off the Hippo pathway. The enhanced AXL expression, in turn, accelerates the metastasis process of HNC by enhancing cell migration, invasion, and vascular/lymphatic angiogenesis. Targeting AXL with R428 in HNC, did not reduce cell proliferation, but weakened vascular/lymphatic angiogenesis, and reversed EMT process, resulting in reduced metastasis. Targeting HRAS with tipifarnib activates the Hippo pathway, enhances YAP1 cytoplasmic retention, and reduces YAP1 nuclear export and AXL expression, thereby reducing HNC metastasis.

Figure 7.

Schematic representation of the molecular mechanism of HRAS-mutation–mediated metastasis in HNC. Top, in the disease state, HRASmut inhibits the Hippo pathway, preventing YAP1 degradation and, in turn, leading to nuclear export of YAP1, thereby regulating the transcriptional expression of multiple genes, including AXL. AXL overexpression and activation by its ligand Gas6 enhance the migratory activity of tumor cells and upregulate the expression of VEGFA, VEGFC, and other angiogenic factors as well as EMT genes. This activation of HRAS–YAP1–AXL enhanced angiogenesis, sustained proliferation, and increased metastasis potential, causing metastatic spread to the cLNs and lungs. Middle, treament of tumors with R428 did not reduce tumor cell proliferation but did reduce the metastasis potential of tumor cells and tumor angiogenesis by downregulating angiogenic factors as well as reversing the EMT, leading to diminished metastatic spread to the cLNs and lungs. Bottom, treament of tumors with tipifarnib blocked the activation of HRAS, thus switching off the Hippo pathway and causing YAP1 degradation and downregulation of AXL expression. This deactivation of HRAS–YAP1–AXL reduced the proliferation and metastasis potential of tumor cells and tumor angiogenesis, leading to diminished metastatic spread to the cLNs and lungs. (Created with BioRender.com.)

Figure 7.

Schematic representation of the molecular mechanism of HRAS-mutation–mediated metastasis in HNC. Top, in the disease state, HRASmut inhibits the Hippo pathway, preventing YAP1 degradation and, in turn, leading to nuclear export of YAP1, thereby regulating the transcriptional expression of multiple genes, including AXL. AXL overexpression and activation by its ligand Gas6 enhance the migratory activity of tumor cells and upregulate the expression of VEGFA, VEGFC, and other angiogenic factors as well as EMT genes. This activation of HRAS–YAP1–AXL enhanced angiogenesis, sustained proliferation, and increased metastasis potential, causing metastatic spread to the cLNs and lungs. Middle, treament of tumors with R428 did not reduce tumor cell proliferation but did reduce the metastasis potential of tumor cells and tumor angiogenesis by downregulating angiogenic factors as well as reversing the EMT, leading to diminished metastatic spread to the cLNs and lungs. Bottom, treament of tumors with tipifarnib blocked the activation of HRAS, thus switching off the Hippo pathway and causing YAP1 degradation and downregulation of AXL expression. This deactivation of HRAS–YAP1–AXL reduced the proliferation and metastasis potential of tumor cells and tumor angiogenesis, leading to diminished metastatic spread to the cLNs and lungs. (Created with BioRender.com.)

Close modal

HNC is characterized by a poor prognosis with a high risk of local recurrence and distant metastasis (1, 24, 25). Here, we showed that mutations in HRAS promoted cancer aggressiveness and metastasis, findings that are in line with the oncogenic effect of RAS family members and their signaling pathways in other cancer types (26–31). Our study provided novel insights into the role(s) of HRASmut in promoting metastasis and identifies the underlying signaling mechanism. Specifically, we found that HRASmut acts through a Hippo–YAP1–AXL axis, which regulates cLN and distant metastasis in HNC.

Metastasis is a complex process that starts when cancer cells acquire the ability to invade adjacent organs and/or migrate through the bloodstream or lymphatic system to distant sites, where they survive and colonize (32). The current study shows that ectopic expression of the HRAS mutation in HNC cells markedly increased their invasive ability both in vitro and in vivo (Figs. 1 and 2), in keeping with previous reports showing the role of HRAS in promoting the EMT (33, 34) and cell invasion in urothelial and gastric cancers (35, 36). A key finding in the current study was the upregulation of AXL expression in HRASmut HNC cells. AXL is a member of the TAM family of receptors (Tyro, AXL, and MER) and is also a mesenchymal marker implicated in cancer cell proliferation, migration, angiogenesis, metastasis, and therapeutic resistance of HNC (37–40). AXL is also expressed in the partial EMT (pEMT), a prerequisite plasticity stage for metastasis that promotes local invasion in HNC (41). Genomic and transcriptomic analyses of the TCGA cohort showed that HNC tumors harboring the HRAS mutation were clustered as mesenchymal/basal subtypes (42), with AXL expression being higher in those clusters. In a previous study in basal-like breast cancer cells, it was shown that HRAS-induced EMT was dependent on the upregulation of vimentin, which mediated invasion and migration, in part, through upregulating AXL (43). In the current study, we showed a different mechanism of regulation of AXL expression by HRASmut, which involves activation of YAP1 and the inhibition of the Hippo pathway. Using HRASmut HNC cell lines, we demonstrated that AXL expression is regulated by YAP1, and we thus extensively studied the role of AXL expression in determining metastasis to the cLNs and lungs in mice. Notably, the levels of phosphorylated AXL were undetectable, so it was not possible to examine the role of AXL activation in the metastasis of HRASmut HNC. The overexpression of AXL in AXL-null HRASmut HNC cells with a low metastatic colonization potential led to augmented migratory and invasive abilities in vivo. The AXL-depleted models showed reduced metastasis, corroborating studies from other AXL-depleted cancer lines (44–46). In the current study, we also demonstrated that AXL depletion blocked neovascularization, thus preventing the movement of tumor cells to draining lymph nodes and the bloodstream. Pharmacological inhibition of AXL reduced the metastatic burden but did not curtail tumor progression. These results indicate that AXL expression in HRASmut HNC is probably a prerequisite for the metastasis process, but these tumor cells are resistant to anti-AXL therapies, as has been shown for resistance to erlotinib (37). Our in vivo data suggest that inhibition of AXL signaling alone is insufficient to eliminate tumors in HRAS-mutant HNC cancers, and more potent treatment strategies are required.

Because AXL is one of the downstream targets of YAP/TAZ (16, 47), the known crosstalk between AXL and YAP1 in promoting tumor progression in HNC (48) motivated us to examine the HRAS–YAP1–AXL axis. RAS can regulate YAP1 activity by multiple mechanisms, namely through: (i) its effector, PI3K (49), (ii) upregulation of RTKs or RTK ligands (50), (iii) SUMOylation of ubiquitin thioesterase (OTUB2; ref. 51), (iv) downregulation of MST1/2 activity by inducing the formation of inactive MST1/MST2 heterodimers (17), and/or (v) downregulation of suppressor of cytokine signaling 6 (SOCS6; ref. 52). YAP1 is a potent transcriptional coactivator that lacks direct DNA-binding activity but modulates gene expression by binding to other transcription factors. As documented previously, phosphorylation of YAP1 at various serine residues have various biological implications on YAP1 activity. For example, phosphorylation of serine at 109 and 127 leads to YAP1 cytoplasmic retention (18, 53) whereas at 381/397 it causes ubiquitination degradation of YAP1, which is dependent on the β-transducin repeat-containing E3 ubiquitin protein ligase complex (18, 54). Our study shows that HRAS signaling regulates YAP1 activity and stability by turning off the Hippo pathway, and that knocking down HRAS or blocking HRAS signaling with tipifarnib turned on the Hippo pathway, enhanced YAP1 sequestration in the cytoplasm and reduced nuclear localization, which subsequently led to reduction in AXL expression.

HRAS proteins undergo several post-translational modifications that facilitate their attachment to membranes, including the addition of a farnesyl isoprenoid moiety by the enzyme farnesyl transferase (55). Farnesyl transferase inhibitors reduce HRAS membrane localization, leading to decreased GTP-bound HRAS and decreased signaling through RAS effector pathways, thus blocking HRAS signaling (56, 57). Among the farnesyl transferase inhibitors, tipifarnib is a potent compound that has demonstrated compelling antitumor activity in preclinical models (23) and in a heavily pretreated cohort of patients with recurrent/metastatic HNC carrying HRAS mutations (58, 59). In this work, we showed for the first time that tipifarnib treatment attenuates the metastatic potential of HRASmut cells, and we also provided mechanistic insight into the effect of tipifarnib on the Hippo pathway, YAP1 activation, and AXL expression. The efficacy of tipifarnib in preventing primary tumor growth was relatively modest in this study with no regression noted. Previous studies have described mechanisms of resistance to tipifarnib, including activation of the MAPK and PI3K pathways (60). Further investigation is required to explore the efficacy of a treatment combination of tipifarnib and PI3K inhibitors in syngeneic murine-HRASV12 shp53 EpT tumors, as it has been shown that a combination of the two compounds exhibited superior antitumor activity in PDXs (61).

The growth of primary tumors and metastatic neoplastic lesions depends strongly on the ability of cancer cells to initiate neovascularization through angiogenesis (62, 63). Tumor cells induce angiogenesis by secreting pro-angiogenic factors, such as VEGFs and angiopoietins, and releasing proteolytic enzymes that recruit endothelial cells and cleave extracellular matrices (63). Angiogenesis is thus an integral aspect of the growth, proliferation, and metastasis of HNC and has potential implications for the prognosis and treatment of both localized and recurrent/metastatic HNC (64). For instance, the upregulation of angiogenic factors, such as VEGFA, VEGFC, and VEGFD, typically corresponds to increased vascularity, lymph node metastasis, an inadequate response to cytotoxic chemotherapy, and advanced disease with a poor prognosis in HNC (as reviewed in ref. 65). It is well documented that oncogenic HRAS functions as a cellular switch that controls angiogenesis and vascular permeability by activating distinct downstream effectors (66, 67). Our study is also in line with previous findings in breast cancer (68) that AXL depletion in HRASmut cells attenuates the expression of the pro-angiogenic factors, VEGFA and VEGFC, thereby reducing vascular and lymphatic angiogenesis. Recent studies on AXL depletion in HER2+ breast cancers reveal that, during hypoxia, interfering with AXL reduces HIF1α levels and AKT phosphorylation, leading to normalizing the blood vessels and thus promotes a proinflammatory microenvironment (45). Our in vivo findings on reduced neovascularization and metastasis on AXL depletion also imply the significance of AXL and its involvement in controlling hypoxia-mediated angiogenesis and the tumorigenesis process. Further studies are required to study the role of hypoxia and AXL in HRASmut HNC.

In summary, our studies revealed that metastasis in HRASmut HNC is regulated by the stabilization of YAP1 activity and the overexpression of AXL, both of which regulate cell migration, invasion, and vascular/lymphatic angiogenesis. Blocking HRAS signaling reduces YAP1 activity and AXL expression, which, in turn, downregulate VEGFA and VEGFC expressions. Ultimately, we provide evidence that targeting the novel HRAS–YAP1–AXL signaling axis may represent an attractive therapeutic candidate to limit the metastasis of HRASmut HNC cells.

G. Gausdal reports support from BerGenBio ASA during the conduct of the study. A.L. Ho reports grants and personal fees from Kura Oncology and Elevar Therapeutics, Physician Education Resources, Affyimmune, Rgenta, Remix, Clinical Endocrinology Update, Merck, ExpertConnect, University of Pittsburgh, New York University, Prelude Therapeutics, Eisai and Ayala Therapeutics, NIH, Exelixis, Emory University, ASTRO, Massachusetts General Hospital, McGiveny Global Advisors, Shanghai Jai Tong University, Bayer, Onc4, Bioatla, Poseida Therapeutics, Novartis, AstraZeneca, Genentech, Celldex, Bristol-Meyer Squibb, Astellas, Inxmed, and Cellestia outside the submitted work; as well as a patent for Lesional Dosimetry Methods for Tailoring Targeted Radiotherapy in Cancer issued; and is the international principal investigator for the HNC registration clinical trial for Kura Oncology. Memorial Sloan Kettering Cancer Center has a filed patent using farnesyl–transferase inhibition to target HRAS mutations in thyroid cancer. A.L. Ho is not a licensee or inventor on that patent. A.J. Rosenberg reports personal fees from Astellas, EMD Serono, Novartis, Eisai, and Nanobiotix and support from Privo and Galectin outside the submitted work. L. Kessler reports employment at Kura Oncology, Inc. F. Burrows reports employment at Kura Oncology, Inc. J.R. Grandis is the co-inventor of a decoy oligonucleotide targeting STAT3 that has been licensed by BluedotBio outside the submitted work. J. Gutkind reports grants from NIH/NCI and nonfinancial support from Kura Pharmaceuticals during the conduct of the study; as well as personal fees from Domain Therapeutics, and Pangea Therapeutics and support from io9 and Kadima Pharmaceuticals, Dracen Pharmaceuticals, Revolution Medicine, and Verastem outside the submitted work. M. Elkabets reports grants from Kura Oncology during the conduct of the study. No disclosures were reported by the other authors.

S. Jagadeeshan: Conceptualization, data curation, formal analysis, validation, investigation, methodology, writing–original draft, writing–review and editing. M. Prasad: Data curation, investigation, methodology. M. Badarni: Investigation. T. Ben-Lulu: Investigation. V.B. Liju: Validation, investigation. S. Mathukkada: Investigation. C. Saunders: Investigation. A.B. Shnerb: Investigation. J. Zorea: Data curation, investigation, methodology. K.M. Yegodayev: Methodology. M. Wainer: Methodology. L. Vtorov: Investigation. I. Allon: Data curation, investigation, methodology. O. Cohen: Resources, data curation, software, validation, methodology. G. Gausdal: Resources. D. Friedmann-Morvinski: Resources. S.C. Cheong: Resources. A.L. Ho: Investigation. A.J. Rosenberg: Resources. L. Kessler: Resources. F. Burrows: Resources. D. Kong: Resources, investigation. J.R. Grandis: Resources, writing–review and editing. J. Gutkind: Resources, writing–review and editing. M. Elkabets: Conceptualization, resources, supervision, funding acquisition, writing–original draft, project administration, writing–review and editing.

This work was funded by the Israel Science Foundation (ISF, 302/21; to M. Elkabets), The United States–Israel Binational Science Foundation (BSF, #2021055 to M. Elkabets and J. Gutkind), ISF and NSFC Israel–China project (to M. Elkabets and D. Kong; #3409/20), and Kura Oncology (to M. Elkabets). At Memorial Sloan Kettering Cancer Center, support was provided by the NIH/NCI Cancer Center support grant P30 CA008748 (to A.L. Ho). S. Jagadeeshan is the recipient of a PBC postdoctoral fellowship from the Israel Council for Higher Education. The authors thank Dr. S. Ovadia, Ben Gurion University of the Negev, for her assistance in the animal facility; Inez and Prof. Neta Erez, Tel Aviv University for the critical editing of the article.

The publication costs of this article were defrayed in part by the payment of publication fees. Therefore, and solely to indicate this fact, this article is hereby marked “advertisement” in accordance with 18 USC section 1734.

Note: Supplementary data for this article are available at Cancer Research Online (http://cancerres.aacrjournals.org/).

1.
Pisani
P
,
Airoldi
M
,
Allais
A
,
Valletti
PA
,
Battista
M
,
Benazzo
M
, et al
.
Metastatic disease in head and neck oncology
.
Acta Otorhinolaryngol Ital
2020
;
40
:
S1
86
.
2.
Beckham
TH
,
Leeman
JE
,
Xie
P
,
Li
X
,
Goldman
DA
,
Zhang
Z
, et al
.
Long-term survival in patients with metastatic head and neck squamous cell carcinoma treated with metastasis-directed therapy
.
Br J Cancer
2019
;
121
:
897
903
.
3.
Shaikh
H
,
Karivedu
V
,
Wise-Draper
TM
.
Managing recurrent metastatic head and neck cancer
.
Hematol Oncol Clin North Am
2021
;
35
:
1009
20
.
4.
Wan
L
,
Pantel
K
,
Kang
Y
.
Tumor metastasis: moving new biological insights into the clinic
.
Nat Med
2013
;
19
:
1450
64
.
5.
Fares
J
,
Fares
MY
,
Khachfe
HH
,
Salhab
HA
,
Fares
Y
.
Molecular principles of metastasis: a hallmark of cancer revisited
.
Signal Transduct Target Ther
2020
;
5
:
1
17
.
6.
Colella
S
,
Richards
KL
,
Bachinski
LL
,
Baggerly
KA
,
Tsavachidis
S
,
Lang
JC
, et al
.
Molecular signatures of metastasis in head and neck cancer
.
Head Neck
2008
;
30
:
1273
83
.
7.
Bhave
SL
,
Teknos
TN
,
Pan
Q
.
Molecular parameters of head and neck cancer metastasis
.
Crit Rev Eukaryot Gene Expr
2011
;
21
:
143
53
.
8.
Novoplansky
O
,
Jagadeeshan
S
,
Regev
O
,
Menashe
I
,
Elkabets
M
,
Segal
S
.
Worldwide prevalence and clinical characteristics of RAS mutations in head and neck cancer: a systematic review and meta-analysis
.
Front Oncol
2022
;
12
:
838911
.
9.
Reder
H
,
Wagner
S
,
Wuerdemann
N
,
Langer
C
,
Sandmann
S
,
Braeuninger
A
, et al
.
Mutation patterns in recurrent and/or metastatic oropharyngeal squamous cell carcinomas in relation to human papillomavirus status
.
Cancer Med
2021
;
10
:
1347
56
.
10.
Braig
F
,
Voigtlaender
M
,
Schieferdecker
A
,
Busch
CJ
,
Laban
S
,
Grob
T
, et al
.
Liquid biopsy monitoring uncovers acquired RAS-mediated resistance to cetuximab in a substantial proportion of patients with head and neck squamous cell carcinoma
.
Oncotarget
2016
;
7
:
42988
95
.
11.
Takács
T
,
Kudlik
G
,
Kurilla
A
,
Szeder
B
,
Buday
L
,
Vas
V
.
The effects of mutant Ras proteins on the cell signalome
.
Cancer Metastasis Rev
2020
;
39
:
1051
65
.
12.
Santra
T
,
Herrero
A
,
Rodriguez
J
,
von Kriegsheim
A
,
Iglesias-Martinez
LF
,
Schwarzl
T
, et al
.
An integrated global analysis of compartmentalized HRAS signaling
.
Cell Rep
2019
;
26
:
3100
15
.
13.
Friedmann-Morvinski
D
,
Bushong
EA
,
Ke
E
,
Soda
Y
,
Marumoto
T
,
Singer
O
, et al
.
Dedifferentiation of neurons and astrocytes by oncogenes can induce gliomas in mice
.
Science
2012
;
338
:
1080
4
.
14.
Carpentier
G
.
Dot blot analyzer: software development using the macro language of ImageJ
.
In: Proceedings of the ImageJ User and Developer Conference. Centre de Recherche Public Henri Tudor: Esch-sur-Alzette, Luxembourg
.
2008
;
Volume 189
, pp.
3-5-189
.
15.
Antony
J
,
Huang
RYJ
.
AXL-driven EMT state as a targetable conduit in cancer
.
Cancer Res
2017
;
77
:
3725
32
.
16.
Xu
MZ
,
Chan
SW
,
Liu
AM
,
Wong
KF
,
Fan
ST
,
Chen
J
, et al
.
AXL receptor kinase is a mediator of YAP-dependent oncogenic functions in hepatocellular carcinoma
.
Oncogene
2011
;
30
:
1229
40
.
17.
Rawat
SJ
,
Araiza-Olivera
D
,
Arias-Romero
LE
,
Villamar-Cruz
O
,
Prudnikova
TY
,
Roder
H
, et al
.
H-ras inhibits the hippo pathway by promoting Mst1/Mst2 heterodimerization
.
Curr Biol
2016
;
26
:
1556
63
.
18.
Zhao
B
,
Li
L
,
Tumaneng
K
,
Wang
CY
,
Guan
KL
.
A coordinated phosphorylation by Lats and CK1 regulates YAP stability through SCFβ-TRCP
.
Genes Dev
2010
;
24
:
72
85
.
19.
Pleasance
E
,
Titmuss
E
,
Williamson
L
,
Kwan
H
,
Culibrk
L
,
Zhao
EY
, et al
.
Pan-cancer analysis of advanced patient tumors reveals interactions between therapy and genomic landscapes
.
Nat Cancer
2020
;
1
:
452
68
.
20.
Boonkitticharoen
V
,
Kulapaditharom
B
,
Leopairut
J
,
Kraiphibul
P
,
Larbcharoensub
N
,
Cheewaruangroj
W
, et al
.
Vascular endothelial growth factor a and proliferation marker in prediction of lymph node metastasis in oral and pharyngeal squamous cell carcinoma
.
Arch Otolaryngol Head Neck Surg
2008
;
134
:
1305
11
.
21.
Cohen-Kaplan
V
,
Naroditsky
I
,
Zetser
A
,
Ilan
N
,
Vlodavsky
I
,
Doweck
I
.
Heparanase induces VEGF C and facilitates tumor lymphangiogenesis
.
Int J Cancer
2008
;
123
:
2566
73
.
22.
Sugiura
T
,
Inoue
Y
,
Matsuki
R
,
Ishii
K
,
Takahashi
M
,
Abe
M
, et al
.
VEGF-C and VEGF-D expression is correlated with lymphatic vessel density and lymph node metastasis in oral squamous cell carcinoma: implications for use as a prognostic marker
.
Int J Oncol
2009
;
34
:
673
80
.
23.
Gilardi
M
,
Wang
Z
,
Proietto
M
,
Chilla
A
,
Calleja-Valera
JL
,
Goto
Y
, et al
.
Tipifarnib as a precision therapy for HRAS-mutant head and neck squamous cell carcinomas
.
Mol Cancer Ther
2020
;
19
:
1784
96
.
24.
Johnson
DE
,
Burtness
B
,
Leemans
CR
,
Lui
VWY
,
Bauman
JE
,
Grandis
JR
.
Head and neck squamous cell carcinoma
.
Nat Rev Dis Primers
2020
;
6
:
1
22
.
25.
Wise-Draper
TM
,
Bahig
H
,
Tonneau
M
,
Karivedu
V
,
Burtness
B
.
Current therapy for metastatic head and neck cancer: evidence, opportunities, and challenges
.
Am Soc Clin Oncol Educ Book
2022
;
42
:
1
14
.
26.
Pei
L
,
Melmed
S
,
Scheithauer
B
,
Kovacs
K
,
Prager
D
.
H-ras mutations in human pituitary carcinoma metastases
.
J Clin Endocrinol Metab
1994
;
78
:
842
6
.
27.
Jakob
JA
,
Bassett
RL
,
Ng
CS
,
Curry
JL
,
Joseph
RW
,
Alvarado
GC
, et al
.
NRAS mutation status is an independent prognostic factor in metastatic melanoma
.
Cancer
2012
;
118
:
4014
23
.
28.
Boutin
AT
,
Liao
WT
,
Wang
M
,
Hwang
SS
,
Karpinets
TV
,
Cheung
H
, et al
.
Oncogenic Kras drives invasion and maintains metastases in colorectal cancer
.
Genes Dev
2017
;
31
:
370
82
.
29.
Yeh
HH
,
Tseng
YF
,
Hsu
YC
,
Lan
SH
,
Wu
SY
,
Raghavaraju
G
, et al
.
Ras induces experimental lung metastasis through upregulation of RbAp46 to suppress RECK promoter activity
.
BMC Cancer
2015
;
15
:
1
14
.
30.
Wright
KL
,
Adams
JR
,
Liu
JC
,
Loch
AJ
,
Wong
RG
,
Jo
CEB
, et al
.
Ras signaling is a key determinant for metastatic dissemination and poor survival of luminal breast cancer patients
.
Cancer Res
2015
;
75
:
4960
72
.
31.
Yaeger
R
,
Cowell
E
,
Chou
JF
,
Gewirtz
AN
,
Borsu
L
,
Vakiani
E
, et al
.
RAS mutations affect pattern of metastatic spread and increase propensity for brain metastasis in colorectal cancer
.
Cancer
2015
;
121
:
1195
203
.
32.
Leong
SP
,
Naxerova
K
,
Keller
L
,
Pantel
K
,
Witte
M
.
Molecular mechanisms of cancer metastasis via the lymphatic versus the blood vessels
.
Clin Exp Metastasis
2022
;
39
:
159
79
.
33.
Oft
M
,
Peli
J
,
Rudaz
C
,
Schwarz
H
,
Beug
H
,
Reichmann
E
.
TGF-β1 and Ha-Ras collaborate in modulating the phenotypic plasticity and invasiveness of epithelial tumor cells
.
Genes Dev
1996
;
10
:
2462
77
.
34.
Oft
M
,
Akhurst
RJ
,
Balmain
A
.
Metastasis is driven by sequential elevation of H-ras and Smad2 levels
.
Nat Cell Biol
2002
;
4
:
487
94
.
35.
Wu
XY
,
Liu
WT
,
Wu
ZF
,
Chen
C
,
Liu
JY
,
Wu
GN
, et al
.
Identification of HRAS as cancer-promoting gene in gastric carcinoma cell aggressiveness
.
Am J Cancer Res
2016
;
6
:
1935
.
36.
He
F
,
Melamed
J
,
Tang
MS
,
Huang
C
,
Wu
XR
.
Oncogenic HRAS activates epithelial-to-mesenchymal transition and confers stemness to p53-deficient urothelial cells to drive muscle invasion of basal subtype carcinomas
.
Cancer Res
2015
;
75
:
2017
28
.
37.
Giles
KM
,
Kalinowski
FC
,
Candy
PA
,
Epis
MR
,
Zhang
PM
,
Redfern
AD
, et al
.
Axl mediates acquired resistance of head and neck cancer cells to the epidermal growth factor receptor inhibitor erlotinib
.
Mol Cancer Ther
2013
;
12
:
2541
58
.
38.
von Mässenhausen
A
,
Brägelmann
J
,
Billig
H
,
Thewes
B
,
Queisser
A
,
Vogel
W
, et al
.
Implication of the receptor tyrosine kinase AXL in head and neck cancer progression
.
Int J Mol Sci
2017
;
18
:
7
.
39.
Brand
TM
,
Iida
M
,
Stein
AP
,
Corrigan
KL
,
Braverman
CM
,
Luthar
N
, et al
.
AXL mediates resistance to cetuximab therapy
.
Cancer Res
2014
;
74
:
5152
64
.
40.
Badarni
M
,
Prasad
M
,
Balaban
N
,
Zorea
J
,
Yegodayev
KM
,
Joshua
BZ
, et al
.
Repression of AXL expression by AP-1/JNK blockage overcomes resistance to PI3Ka therapy
.
JCI Insight
2019
;
5
:
e125341
.
41.
Puram
SV
,
Parikh
AS
,
Tirosh
I
.
Single cell RNA-seq highlights a role for a partial EMT in head and neck cancer
.
Mol Cell Oncol
2018
;
5
:
01
6
.
42.
Lawrence
MS
,
Sougnez
C
,
Lichtenstein
L
,
Cibulskis
K
,
Lander
E
,
Gabriel
SB
, et al
.
Comprehensive genomic characterization of head and neck squamous cell carcinomas
.
Nature
2015
;
517
:
576
82
.
43.
Vuoriluoto
K
,
Haugen
H
,
Kiviluoto
S
,
Mpindi
JP
,
Nevo
J
,
Gjerdrum
C
, et al
.
Vimentin regulates EMT induction by Slug and oncogenic H-Ras and migration by governing Axl expression in breast cancer
.
Oncogene
2011
;
30
:
1436
48
.
44.
Goyette
MA
,
Duhamel
S
,
Aubert
L
,
Pelletier
A
,
Savage
P
,
Thibault
MP
, et al
.
The receptor tyrosine kinase AXL is required at multiple steps of the metastatic cascade during HER2-positive breast cancer progression
.
Cell Rep
2018
;
23
:
1476
90
.
45.
Goyette
MA
,
Elkholi
IE
,
Apcher
C
,
Kuasne
H
,
Rothlin
CV
,
Muller
WJ
, et al
.
Targeting Axl favors an antitumorigenic microenvironment that enhances immunotherapy responses by decreasing Hif-1α levels
.
Proc Natl Acad Sci U S A
2021
;
118
:
e2023868118
.
46.
Du
W
,
Phinney
NZ
,
Huang
H
,
Wang
Z
,
Westcott
J
,
Toombs
JE
, et al
.
AXL is a key factor for cell plasticity and promotes metastasis in Pancreatic cancer
.
Mol Cancer Res
2021
;
19
:
1412
21
.
47.
Ghiso
E
,
Migliore
C
,
Ciciriello
V
,
Morando
E
,
Petrelli
A
,
Corso
S
, et al
.
YAP-dependent AXL overexpression mediates resistance to EGFR inhibitors in NSCLC
.
Neoplasia
2017
;
19
:
1012
21
.
48.
Li
J
,
Shi
C
,
Zhou
R
,
Han
Y
,
Xu
S
,
Ma
H
, et al
.
The crosstalk between AXL and YAP promotes tumor progression through STAT3 activation in head and neck squamous cell carcinoma
.
Cancer Sci
2020
;
111
:
3222
35
.
49.
Basu
S
,
Totty
NF
,
Irwin
MS
,
Sudol
M
,
Downward
J
.
Akt phosphorylates the Yes-associated protein, YAP, to induce interaction with 14–3-3 and attenuation of p73-mediated apoptosis
.
Mol Cell
2003
;
11
:
11
23
.
50.
Zhang
J
,
Ji
JY
,
Yu
M
,
Overholtzer
M
,
Smolen
GA
,
Wang
R
, et al
.
YAP-dependent induction of amphiregulin identifies a non–cell-autonomous component of the Hippo pathway
.
Nat Cell Biol
2009
;
11
:
1444
50
.
51.
Zhang
Z
,
Du
J
,
Wang
S
,
Shao
L
,
Jin
K
,
Li
F
, et al
.
OTUB2 promotes cancer metastasis via hippo-independent activation of YAP and TAZ
.
Mol Cell
2019
;
73
:
7
21
.
52.
Hong
X
,
Nguyen
HT
,
Chen
Q
,
Zhang
R
,
Hagman
Z
,
Voorhoeve
PM
, et al
.
Opposing activities of the Ras and Hippo pathways converge on regulation of YAP protein turnover
.
EMBO J
2014
;
33
:
2447
57
.
53.
Zhao
B
,
Wei
X
,
Li
W
,
Udan
RS
,
Yang
Q
,
Kim
J
, et al
.
Inactivation of YAP oncoprotein by the Hippo pathway is involved in cell contact inhibition and tissue growth control
.
Genes Dev
2007
;
21
:
2747
61
.
54.
Dong
J
,
Feldmann
G
,
Huang
J
,
Wu
S
,
Zhang
N
,
Comerford
SA
, et al
.
Elucidation of a universal size-control mechanism in drosophila and mammals
.
Cell
2007
;
130
:
1120
33
.
55.
Rowinsky
EK
,
Windle
JJ
,
Von Hoff
DD
.
Ras protein farnesyltransferase: a strategic target for anticancer therapeutic development
.
J Clin Oncol
1999
;
17
:
3631
52
.
56.
Kohl
NE
,
Omer
CA
,
Davide
JP
,
Hamilton
K
,
Koblan
KS
,
Kral
AM
, et al
.
Inhibition of farnesyltransferase induces regression of mammary and salivary carcinomas in ras transgenic mice
.
Nat Med
1995
;
1
:
792
7
.
57.
End
DW
,
Smets
G
,
Todd
AV
,
Applegate
TL
,
Fuery
CJ
,
Patrick
A
, et al
.
Characterization of the antitumor effects of the selective farnesyl protein transferase inhibitor R115777 in vivo and in vitro
.
Cancer Res
2001
;
61
:
131
7
.
58.
Hanna
GJ
,
Guenette
JP
,
Chau
NG
,
Sayehli
CM
,
Wilhelm
C
,
Metcalf
R
, et al
.
Tipifarnib in recurrent, metastatic HRAS-mutant salivary gland cancer
.
Cancer
2020
;
126
:
3972
81
.
59.
Ho
AL
,
Brana
I
,
Haddad
R
,
Bauman
J
,
Bible
K
,
Oosting
S
, et al
.
Tipifarnib in head and neck squamous cell carcinoma with HRAS mutations
.
J Clin Oncol
2021
;
39
:
1856
64
.
60.
Javaid
S
,
Schaefer
A
,
Goodwin
CM
,
Nguyen
VV
,
Massey
FL
,
Pierobon
M
, et al
.
Concurrent inhibition of ERK and Farnesyltransferase suppresses the growth of HRAS mutant head and neck squamous cell carcinoma
.
Mol Cancer Ther
2022
;
21
:
762
74
.
61.
Smith
A
,
Chan
S
,
McCloskey
A
,
Vora
H
,
Burrows
F
,
Malik
S
.
Antitumor activity of Tipifarnib and PI3K pathway inhibitor in HRAS-associated HNSCC
.
Int J Radi Oncol
2022
;
112
:
e48
.
62.
Zetter
BR
.
Angiogenesis and tumor metastasis
.
Annu Rev Med
1998
;
49
:
407
24
.
63.
Bielenberg
DR
,
Zetter
BR
.
The contribution of angiogenesis to the process of metastasis
.
Cancer J
2015
;
21
:
267
73
.
64.
Micaily
I
,
Johnson
J
,
Argiris
A
.
An update on angiogenesis targeting in head and neck squamous cell carcinoma
.
Cancers Head Neck
2020
;
5
:
5
.
65.
Carla
C
,
Daris
F
,
Cecilia
B
,
Francesca
B
,
Francesca
C
,
Paolo
F
.
Angiogenesis in head and neck cancer: a review of the literature
.
J Oncol
2012
;
2012
:
358472
.
66.
Serban
D
,
Leng
J
,
Cheresh
D
.
H-ras regulates angiogenesis and vascular permeability by activation of distinct downstream effectors
.
Circ Res
2008
;
102
:
1350
8
.
67.
Arbiser
JL
,
Moses
MA
,
Fernandez
CA
,
Ghiso
N
,
Cao
Y
,
Klauber
N
, et al
.
Oncogenic H-ras stimulates tumor angiogenesis by two distinct pathways
.
Proc Natl Acad Sci U S A
1997
;
94
:
861
6
.
68.
Tanaka
M
,
Siemann
DW
.
Axl signaling is an important mediator of tumor angiogenesis
.
Oncotarget
2019
;
10
:
2887
98
.